|
Assessing the Impact of Changes in Extreme Precipitation Events on Shallow Landslide Abundance, Location, and Size
Dino
Bellugi, MIT, dinob@mit.edu
(Presenter)
Catherine
Slesnick, Draper Laboratory, cslesnick@draper.com
Erin
Leidy, Draper Laboratory, eleidy@draper.com
Natalya
Markuzon, Draper Laboratory, nmarkuzon@draper.com
Paul
A
O’Gorman, MIT, pog@mit.edu
Taylor
J.
Perron, MIT, perron@mit.edu
John
Regan, Draper Laboratory, jregan@draper.com
Adam
Schlosser, MIT, casch@mit.edu
John
West, Draper Laboratory, jwest@draper.com
Shallow landslides are a widespread phenomenon in the United States and the world. Often triggered by extreme precipitation events, they can be the primary sources of debris flows, and are generally a threatening source of hazards, causing loss of life, destruction of property and infrastructure, and affecting communities all across the nation. It is crucial to accurately assess such hazards, particularly in light of expected climate and land use changes.
Here we explore the landsliding response of a prototype landscape located in the Oregon Coast Range (OCR) to hypothetical changes in intensity, duration, and frequency of extreme rainfall events. We adopt a mechanistic landslide prediction procedure which couples a three-dimensional slope stability model with an efficient search algorithm to predict discrete shallow landslides. We use a landslide inventory collected by repeat field mapping over a 10-year period in an area with constraints on soil, vegetation, hydrological, and rainfall characteristics. In hindcast mode, the procedure reproduces the distribution of sizes and locations of the landslide inventory under a suite of rainfall and moisture characteristics representative of the observation period. We use projections of precipitation extremes under different climate change scenarios to generate landslide forecasts and explore the sensitivity of landslide abundance, size and location to the intensity, duration, and frequency of rainfall events, as well as to antecedent moisture conditions, resulting from the different scenarios.
We also present progress in the development of a data-driven approach to understanding landslide activity and the response to changes in extreme precipitation in an evolving climate. Resultant models forecast landslides based on a combination of remote sensing data and historical surface observations including weather patterns, landcover and lithology, and topographic attributes. We present spatially explicit results from the application of a non-linear classifier (a support vector machine constructed using topographic, soil, and vegetation attributes) to the OCR dataset and compare the results to the existing landslide inventory. Finally, we address the need for eco-atmo-geo-hydrological models to capture the linkages between climate, vegetation, and the landscape.
Presentation:
2013_Poster_Bellugi_19_97.pdf (16390k)
Presentation Type: Poster
Session: Poster Session 1-B
(Tue 4:30 PM)
Associated Project(s):
Poster Location ID: 19
|